Here’s the official description of this panel:
Mining the Buzz: Capturing and Analyzing Movie and Brand Data
Companies and individuals are relying on Twitter to build their brands via engagement and/or community support. Who is doing it right, how are they doing it, and what gains are they seeing? Our panel of gurus will discuss how they and their partners are taking advantage of Twitter to raise awareness and build positive perceptions.
2:45pm – 3:10pm / Speakers: Moderator: Shira Lazar, Steve Broback, Andrew Wallenstein [note: after they published the info on the panelists, a marketer from Avatar joined, but I didn't catch his name.]
Shira Lazar, Media Empress from CBS News, moderated the 2nd panel at the Tweethouse.
Andrew Wallstein from the Hollywood Reporter kicked off the accountability discussion. He posed the question to Steve Broback, one of the organizers of the Tweethouse. Andrew asked, “I want to know more about the predictive model…do you see buzz from a film before it’s release date? Can you really make predictions on how a movie will do 6 months or a year before hand, based on how much noise it’s getting on Twitter? I’m dubious that that kind of evidence is there, but maybe your data says otherwise.”
The marketer who ran point on the social media marketing for Avatar said that they started the marketing effort in August ‘09 “way back before the Dec opening. And: the initial buzz wasn’t as positive as what everyone expected to be!”
Steve Broback then answered Wallstein’s question. “If you’re trying to predict movie grosses, you have to look at it in phases. It’s announced – there’s a surge. The trailer appears – there’s a surge. And then they release previews. Now this is the one we really want to tap into – people can see it on YouTube, Twitter, the film’s site…that’s the key turning point where you need to measure the quality of the buzz and sharing. And, of course, then the movie comes out.”
Wallstein challenged him. “But wouldn’t you want to predict the box office based on the Tweets, to see people’s reaction and use that to see how it’s going to do?”
Broback: “Well, after the movie’s announced, you can analyze the buzz and it’s predictive, but only to a point.”
The Avatar marketer returned, “But I’m curious too. Wouldn’t there be some correlation as to what the gross/screen (per screen average) is – so you can tell the theatre owners and distributors how big it’s going to be?”
Shira then interjected, “Paranormal Activity used Twitter and saw incredible results last year. And then on the other side of the spectrum we saw Ashton Kusher put no traditional money into his movie, and he was hoping to only use Social Media, and it tanked. So what’s that saying about the mix with online and offline? And how you can use online to drive offline results?”
The Avatar marketer said the quality of the buzz is driven by the quality of the movie. “If you have a really good movie, social media can help accelerate the distribution because all the retweeting and fan updates will be spreading the positive message for you, saying – ‘It’s a really great movie.’ But if you have a bad movie, it’s going to backfire. Social media in that case can really hurt you…you can’t stop the word of mouth.”
He then asked Steve how you mitigate the negative effects of early bad buzz, “and help turn the tide?”
Broback said, “Well with the opening night, the die is cast – there isn’t a lot you can do. I think you can take what we’re learning right now and starting to apply it to TV shows. Series can change. It’s not too hard to take insights from the TV series feedback [that you're getting on social networks] to impact future episodes – more of one character, less of another.”
“However,” he added “There tends to be a high correlation between Twittering (sheer volume) and high box office.”
Shira interjected, “So it’s quantity over quality.”
Steve confirmed, “High volume is key. But ‘Can hardly wait to see’ is the key indicator with our data…which let’s you see trends on anything and over a variety of timeframes. The other thing that’s interesting is that the demo of the followers is really important to effecting the outcome. If it skews younger and male, they [the fims tend to] do better financially then if skews older and/or female. “Our hypothesis is that men can convince their spouses or girlfriends to go see a movie that’s targeting them…but women can’t do the same with men.”
That observation garnered groans and laughs from the audience. But I suspect it’s true.
The Avatar marketer continued, “We use several tools. We use all of them. Trender, Visible Measures, NRG. We have every research tool known to man. Conversation-listening tools are only effective if you can use a predictive model to influence a campaign. The one gap that I’m seeing is that engagement factor. It’s not just a view. What’s the correlation between a piece of content that goes our and what the real impact on the campaign? As a studio, we’re very protective about how our content is being displayed; how do we translate a 60-foot screen experience… 3D IMAX … into a 320 x 460 pixels YouTube asset? We’re constantly looking at the conversation about releasing content, what type of content and when…and how it impacts the ultimate success of the campaign.”
Shira asked him a follow up, to describe the social media strategy for Avatar.
“We started our social media campaign around our ComiCon preview. That was the start of it – facebook pages, twitter pages, teaser pages for the dot come…all launched at once, in August 2009. And from then on, each milestone of the campaign it was based on new things being released, while trying to maintain a constant stream of content – how revolutionary the 3D technology was, a new article on Jim Cameron, etc. We programmed it like a TV show. Now we’re at 2million fans on Facebook, and we only got there because we’ve been constantly programming it with content. But it is also a conversation, in addition to feeding content into the stream. So we have to monitor the buzz, and respond. ‘Hey @HWR…we love what you said about the materials, we love that scene too!’ Getting into the conversation and really responding to writers, and fans…all of that is important.”
Matt @HWR reminded them of the earlier negative buzz that Avatar had, ahead of the release, and asked how they turned the conversation around.
The marketer from Avatar confessed, “We don’t know. I’ll be honest. We’re still trying to figure out what the right conversation is to be having. Most people don’t really associate studios with customer service, but that’s essentially what it is. Understanding what the feedback means, and going back into the model – giving them additional content that can help influence/improve their opinion of what they think the movie will be like. It’s definitely a conversation.”
Broback turned the conversation back to conversation-monitoring and tracking on Twitter. “There are so many interesting tools out there…and many are free. Trendistic, People Browser, Social Ping – beta service that’s pretty impressive…There’s also Cloud.li which let’s you do Twitter searches and looks for popular terms [related to your search] and creates a cloud out of them. When you talk about creating content, it lets you see trends emerge – you jump on them in realtime and can create content that’s part of a realtime trend.”
He continued, “Searchtastic let’s you do searches on specific Twitter feeds. You can look at my feed and answer the question, ‘Hey does Broback like Wine?’ It’s a little bit sneaky – but let’s say you want to email me but you don’t have my address. You can see if I’ve ever tweeted it ‘Hey email me…’ and posted it in my feed.”
Shira interjected by quoting @joshrose with his question from the #buzzsx feed: “Twitter seems less predictive than reflective. Predictive would mean you could actually dial up Twitter and increase tix sales, no?”
Steve tried to argue with his definition of ‘predictive’ -and then requested that they take it offline.
@HWR’s Andrew Wallstein finished the conversation with this observation: “What about IM— isn’t really that where the action is? The one-to-one vs. one-to-many. That seems to be much more powerful. But so far there’s no way of even tracking that stuff.”
Related posts:
Copyright © 2010. All rights reserved.
Comments on this entry are closed.